AAAI

FedGRPO: Privately Optimizing Foundation Models with Group-Relative Rewards from Domain Client

FedGRPO: 通过领域客户端的群体相对奖励私有优化基础模型

Gongxi Zhu, Hanlin Gu, Lixin Fan, Qiang Yang, Yuxing Han

AI总结 FedGRPO通过群体相对奖励机制，在保护隐私的前提下提升基础模型的性能和通信效率。

Comments Accepted by AAAI 2026 as Oral

ChaosBench-Logic: A Benchmark for Logical and Symbolic Reasoning on Chaotic Dynamical Systems

ChaosBench-Logic：逻辑和符号推理在混沌动力系统上的基准

Noel Thomas

机构 * Noel Thomas 纠错

AI总结 ChaosBench-Logic通过30种混沌动力系统评估LLM逻辑和符号推理能力，揭示其在复杂推理任务中的局限性。

Comments 7 pages, 0 figures , Accepted to AAAI-26 Bridge Program: Logical and Symbolic Reasoning in Language Models (camera-ready)

Journal ref AAAI 2026 Bridge Program on Logical and Symbolic Reasoning in Language Models, Singapore, Jan 2026

TG-Field: Geometry-Aware Radiative Gaussian Fields for Tomographic Reconstruction

TG-Field：基于几何的辐射高斯场用于断层扫描重建

Yuxiang Zhong, Jun Wei, Chaoqi Chen, Senyou An, Hui Huang

AI总结 TG-Field通过几何感知高斯变形框架，提升静态和动态CT重建的精度与效率，尤其在稀疏投影条件下表现优异。

Comments Accepted to AAAI 2026. Project page: https://vcc.tech/research/2026/TG-Field

Vascular anatomy-aware self-supervised pre-training for X-ray angiogram analysis

基于血管解剖的自监督预训练方法用于X射线造影分析

De-Xing Huang, Chaohui Yu, Xiao-Hu Zhou, Tian-Yu Xiang, Qin-Yi Zhang, Mei-Jiang Gui, Rui-Ze Ma, Chen-Yu Wang, Nu-Fang Xiao, Fan Wang, Zeng-Guang Hou

机构 * State Key Laboratory of Multimodal Artificial Intelligence Systems, Institute of Automation, Chinese Academy of Sciences（多模态人工智能系统国家重点实验室，自动化研究所，中国科学院） ； School of Artificial Intelligence, University of Chinese Academy of Sciences（人工智能学院，中国科学院大学） ； Joint Laboratory of Intelligence Science and Technology, Institute of Systems Engineering, Macau University of Science and Technology（智能科学与技术联合实验室，系统工程研究所，澳门科学大学） ； DAMO Academy, Alibaba Group（达摩院，阿里巴巴集团） 纠错

AI总结 本文提出VasoMIM框架，结合解剖引导掩码策略和一致性损失，用于X射线造影分析，通过XA-170K数据集实现高效预训练并取得最佳性能。

Comments 10 pages, 10 figures, 10 tables. Journal version of VasoMIM (AAAI 2026)

Transferable Backdoor Attacks for Code Models via Sharpness-Aware Adversarial Perturbation

通过尖锐性感知对抗扰动实现代码模型的可迁移后门攻击

Shuyu Chang, Haiping Huang, Yanjun Zhang, Yujin Huang, Fu Xiao, Leo Yu Zhang

AI总结 STAB通过尖锐性感知对抗扰动实现代码模型的可迁移后门攻击，提升攻击的可转移性和隐蔽性。

Comments 9 pages, 5 figures, Accepted at AAAI 2026

Self-Adaptive Graph Mixture of Models

自适应图模型混合

Mohit Meena, Yash Punjabi, Abhishek A, Vishal Sharma, Mahesh Chandran

机构 * Fujitsu Research of India（日本富士通印度研究） 纠错

AI总结 SAGMM通过自适应选择和组合不同GNN模型，提升图学习任务的性能和适应性。

Comments Accepted by AAAI 2026

Parallelism Meets Adaptiveness: Scalable Documents Understanding in Multi-Agent LLM Systems

并行与适应性：多智能体大语言模型系统中的可扩展文档理解

Chengxuan Xia, Qianye Wu, Sixuan Tian, Yilun Hao

机构 * University of California, Santa Cruz（加州大学圣克ruz分校） ； Carnegie Mellon University（卡内基梅隆大学） 纠错

AI总结 本文提出一种多智能体大语言模型系统协调框架，通过动态任务路由、双向反馈和并行评估机制，提升文档理解的适应性和效率。

Comments Accepted at AAAI 2026 Workshop on WoMAPF, Camera ready version